desktop strategy

[Daniel Robbins]

Hi All,

So you all know that I have been a bit grumpy lately, about the bug tracker bugs and such.

I feel a bit better now. I wanted to talk about desktop strategy. This relates to servers too. It is really about the priorities of desktop and server.

Funtoo Linux and Linux in general is awesome for servers, compared to pretty much all other operating systems. 

In contrast, Linux is not that great on the desktop, and will more or less suck compared to other operating systems. Ubuntu is what you install on your relatives' computers when you do not want to hear from them in a while, but you don't do it out of love.

I know some of you are hardcore Linux desktop people. That's great - I am happy for you. Just like if you wanted to shave your eyebrows off, I'd be happy for you and your decision. It doesn't mean it looks good.

Some of you make your desktops look really pretty. That's great too. They still don't look as good as MacOS. That doesn't mean MacOS is the ultimate OS, they just actually hire real UI designers and artists to develop the look and feel of their operating system. And Microsoft throws millions towards design and usability studies so that things work well for a huge number of people, even if it isn't "hardcore".

My point is this -- I understand people who are excited about Linux on the desktop. But let's be real. It's not what Linux is really good at.

So what does this mean for Funtoo? It means this -- bumping core packages and potentially creating problems for servers (awesome) to make the desktop (somewhere between crap and marginal) work better is like jumping off a cliff to catch a frisbee. I want you to understand this. 

Read it again. It is like jumping off a cliff to catch a frisbee. It is not smart. It is not a good direction for your life. If you do it to Funtoo, you are throwing me off a cliff to catch a frisbee that you are throwing in my general direction, which I don't appreciate.

What this means for you is that if you are going to be creating bugs saying "we need x, y, z because many users are experiencing problems" and being vague about the reason why, I will mark your bug as rejected. 

Because more than likely, what you really wanted to write but were afraid to say was "we need x, y, z because I love the Linux desktop and want to screw over the people using Funtoo on servers so I can run the latest version of GNOME on my laptop". I am not stupid. I can figure this shit out.

Where does this leave us? With some important take-home points:

1) We will be pursuing a desktop strategy. It WILL NOT be pursued at the expense of server stability, PERIOD.

2) To do this properly, we are going to need the funtoo-1.0 profile to be completed FIRST. BEFORE WE THINK ABOUT DESKTOP STUFF.

3) The desktop will still be a secondary priority to server, but we will still be pursuing it fairly aggressively, but WITHIN REASON.

4) The existing desktop experience on Linux is at best just "OK", so we set our expectations accordingly. Goals are to make the default experience "OK" instead of "crap" and an optimized experience "wow, that's not bad" as opposed to "eh, I just puked a little bit".

5) Once the funtoo-1.0 profile is ready, THEN we will be adding very recent versions of some things and using masking to enable them on Desktop profiles.

6) We are not just going to push new versions of packages due to pressure from Gentoo, upstream, wherever, until funtoo-1.0 profile is ready to go and we can do so safely.

7) There will always be exceptions to this rule if it is time to do a periodic version update or we can do a version bump with no down-side for server. 

8) Bumping a version is a solution to a problem. The bug describes the problem. The problem needs to be described in detail FIRST. Then we investigate the root cause of the problem. This involves research, done in comments. Then we determine the course of action, and BDFL has final say. 

The solution can be a version bump, or it can be something else. If your bug description says "bump version of x", then it will be rejected. Because that's a course of action, not a bug. Repeat this 100 times, make an mp3 of your mom saying it in a sweet voice and listen to it every night before going to bed. If I see another bug that says "bump version x of y" I am going to take a plane to where you live and punch you in the face.

I hope this clarifies things for everyone.

Best Regards,

Daniel

[Marc "Whitewolf Fox" Richter]

Hi Daniel,

I'm not a Funtoo or Linux pro of that level you and the core devs are,
but I understand your points and it is good to get a little more
insight into the directions the distribution heads to and the reasons
for it by reading statements like these.
Although I do not agree to your Linux desktop estimation, I very much
support the "servers first" - strategy you are enforcing to Funtoo!
Since you mentioned the bad word with "U" first: I think that the guys
at Ubuntu haven't focused to their server support enough, even not in
the server edition of Ubuntu. So I'm very happy to read that you
enforce this very important point to Funtoo Linux!

Since I'm a little bit in fear now about a face punch and in general
want to do things right: I have to say a thing from a non-expert sight
you other pro's might be interested in. I think it still is a little
bit difficult to get in touch with helping to develop the distribution
currently. An Funtoo-Bugtracker was indeed needed , but the Redmine
Tracker has been inserted in the projects workflow a little fast and -
as far I see - came with quite less documentation about how and when
to use it. It is, for example, a little bit difficult for a newbee to
see if a bug is better kept there or in the Gentoo bugtracker. Also, I
haven't read a real usage guide about how the Funtoo - Projekt handles
bugs and the tracker. For example: The was / is a problem with a
package I wrote myself in flora once. Since I couldn't figure the
exact solution out in instant, I added a bug to Redmine and set it to
be worked on by myself until I was capable of fixing it, so that
others who might face this problem see, that it is already known and
it's worked on. This, somehow was done with only the best intention,
but Oleg understood it like I push work which should be done by me to
others. I couldn't understand that exactly, since I assigned the bug
to myself.
This seems as if you and the other devs have another kind of working
with a bugtracker than I know, which neither is better or worse, but
different. Perhaps you or someone else close to Funtoos internal
workflows should write a little guide about best practices about how
and when to use the tracker. Your 8 points in this mail are a good
start and example for "Do"'s and "Don't"'s .

Best Regards,
Marc

[Daniel Robbins]

Well, I don't know if there is fantastic documentation on Redmine, and I am personally struggling with it. I may switch to a free version of JIRA as I think it could work better for us.

I see the same types of problems with bugs that you see, and ideally the bug tracker would help us with the workflow a bit more.

-Daniel

--
To manage your subscription, visit this group at
http://groups.google.com/group/funtoo-dev?hl=en
---
Also be sure to check out:
 Funtoo Forums: http://forums.funtoo.org
 Planet Larry: http://larrythecow.org

[Sylvain Alain]

Hi everyone, I understand that right now Funtoo focus on the Server side, but there's a lot of Users that runs Funtoo Current because at the very beginning(back in 2009 if I'm right), Funtoo was there to protect Funtoo Current from all the messy stuff from Gentoo Testing.

 

I agree that right now, it's not possible to update all the stuff for the Funtoo Desktop Users, since Funtoo Current is used by both(Server and Desktop).

 

As soon as the Funtoo profile hit the production stage, I think that 2 separates profiles will be the ultimate solution to handle both concepts.

 

Profile Server : All the stuff that is not related with Desktop and all ultra-stable stuff as usual.

 

Profile Desktop : Try to focus on keeping the pace with Gentoo upstream and have a Funtoo Current rocksolid with the custom ebuilds(udev,openrc,consolekit,coreboot,boot-update etc...).

 

I assume that the Funtoo profile is the tool that can handle multiples versions of packages at the same time, so basically the  Server profile can use the udev-171 version while the Desktop profile use the udev-182.

 

With that kind of mecanic, I think that Funtoo will be able to please the Server gouru and the Desktop fanatics too.

 

Do you think this vision can be possible ?

 

Also, maybe some Coredev will be please to focus on the Server or on the Desktop side...depending of their interest.

 

It's my 2 cents.

 

Sylvain

 

 

2012/5/23 Daniel Robbins <drob...@funtoo.org>

--
Salut
alp
Sylvain

[PerfMonk]

The Funtoo site is pretending this:
=========================================================================================================================================================================================================================
The Gentoo Ecosystem
Our Core Team is focused on advancing the state-of-the-art in Linux
distributions by developing our own improvements to Gentoo Linux,
while remaining compatible with the upstream changes from the Gentoo
Linux project.

We are committed to maintaining high-levels of compatibility and
collaboration with the Gentoo Linux project, and challenge ourselves
to innovate ... etc.
=========================================================================================================================================================================================================================

I see the following words "state-of-the-art" and "challenge ourselves
to innovate"... This is exactly what brought me and many to Funtoo.
A very up to date and clean desktop with everything that comes with
it.
I personnally believe that the linux desktop is mature and do not like
the "we can't win this battle" attitude. The linux desktop does not
sucks at all, I do not agree here.

We should be able to maintain both a server and a desktoop profile
with differents objectives. But I'm still entangled on how many
servers are using Funtoo ?
Why is it so important ? Why make the desktop users second class
citizen ?

We would understand strategic decisions when supported by facts. But
I have seen no facts, just decisions that I don't necessarly endorse.

I'm desappointed bye these choices and I'm no different from many
desktop users that came to Funtoo for innovation.

Sorry to disagree but it must be said, I don't really care about the
server part...

Bernard Tremblay
(?still) a Funtoo user

[Daniel Robbins]

Facts:

Linux has 1% desktop market share. (http://www.omgubuntu.co.uk/2012/01/is-linux-marketshare-on-the-rise-it-seems-so/) - even after ~8 years of Ubuntu.

Linux powers Amazon, Google, etc. (millions of servers)

Linux powers 63% of all Web sites.

Linux powers basically all supercomputers.

Linux powers all Android mobile devices (up to 72% smartphone marketshare in parts of Europe.)

We are not giving up on desktop but we must have priorities and the priorities are "server" (ie. cloud, Web, etc.), followed by mobile/embedded, and desktop is more of a strategic target which we are starting.

If the Linux desktop is a success, I do not think it will be running KDE and GNOME. 

But in order to "win" the desktop, we must live in reality first. And then we can make honest decisions.

For KDE, GNOME, and other DEs/WMs, we will make the experience better. We will be doing some stuff that no one else is doing on the desktop, which will be innovative. 

The first step is to deal with reality. If the Linux desktop is so good, why aren't more people using it? The reason is, even though you may like it, it is not competitive with MacOS or Windows in the larger world for a number of reasons.

-Daniel

[§ĥřïñïďĥï Ŗäö]

well. i use funtoo at my workplace . i wanted something that was rock solid and had all the fun tools of gentoo . so recently i upgraded our office desktops/some server to funtoo . and its working rock solid :)   . i never had a problem from the desktop side or the server side. both work on the same base with  some diff minor kernel tweakings. i get the latest on desktop . i use kde4.8.3 , pulseaudio2 . latest gimp2.8 :) . i have the choice to unmask the live builds too . what more do i need in a bleeding edge desktop?! . well.. i don't have any comments on servers. they just work!. even gentoo just worked!. i think we need to concentrate on stabilizing the base system and keep keeping the upper layer packages uptodate with the latest or can have profiles to control the flow of packages. i agree with mr drobbins on stabilizing the profile system so we can play with it anyhow we want once its done . 

--
To manage your subscription, visit this group at
http://groups.google.com/group/funtoo-dev?hl=en
---
Also be sure to check out:
 Funtoo Forums: http://forums.funtoo.org
 Planet Larry: http://larrythecow.org

--

regards
- shrinidhi


Ęvęņ ģóđ fąįļş ŧŏ ųŋđęŗşţąņđ å ĥųmąņ ųņţĭļ ĥĭş đęąţĥ

http://www.linkedin.com/in/shrinidhi666
http://www.shrinidhi666.wordpress.com
http://www.imdb.com/name/nm3025616

[Adam Smith]

On Wed, May 23, 2012 at 11:26 AM, PerfMonk <tremblay...@gmail.com> wrote:

I personnally believe that the linux desktop is mature and do not like
the "we can't win this battle" attitude.  The linux desktop does not
sucks at all, I do not agree here.

I think Bryan Lunduke always seems to make a good argument on why desktop linux sucks. This talk is from 

LinuxFest Northwest 2012.

http://www.youtube.com/watch?v=Sh-cnaJoGCw

Adam

[Nick Gotsinas]

Hello,

Since my take away from Daniel's email was more priorities rather then
choices I would like to offer my insight and my experiences using
Gentoo/funtoo as servers since 2001.

I am not even going to debate the desktop argument because it makes no
sense, this was not an all or nothing, it was just stating priorities.

Even though the facts Daniel posted are correct it is important to
note that, in these cases, Linux is not Gentoo/funtoo or any source
based info. It is most probably Redhat/CentOS.

In my dealings with almost every IT department I have encountered a
lot of resistance when recommending source based OS's. I mostly
recommended Gentoo until I switched to funtoo in 2009.

In the Gentoo days it was a bit easier for them to accept it because
support was available in some form or another. Support is not even in
the horizon right now but something to consider going forward.
Managers will not sign off on something that only one person can
fix/maintain.

Source based OS's have a lot of issues with updates, nobody wants to
troubleshoot why an ebuild breaks during an update with a small
maintenance window on a production environment, a solution proposed
(by a redhat engineer) was that if funtoo could have an option of
being package based as well it may receive a warmer welcome. If this
already exists I apologize. However if it does this information is not
readily accessible or available.

Lack of available AMI for Funtoo on AWS, Building AMI's is tedious and
long, if a Funtoo official AMI was readily available it would get used
more and thus help growth in the server market. For this reason alone,
my current employer, forces me to use official ubuntu's official
AMI's citing "security" as their reasoning. I bring up the cloud
because in today's reality almost everything starts on the cloud in
some way shape or form.

These are just a few points of many have been the final arguments
brought against my proposals every time.

So to summarize, unless these points are addressed with funtoo I don't
see the reality changing much in the corporate sector or funtoo's
future as a viable server option.

These are not necessarily my opinions, but more my experiences when
faced with this in the past. At least I am allowed to use funtoo on my
workstation and research servers which is nice :)

As always if there is anything I can do to help the project as a
sysadmin I'm willing and available.

Thanks,
Nick

--
<da...@gameforce.ca>
http://www.gameforce.ca

[Peter Panther]

Hi Daniel,

i am quit the same opinion like perlmonk. What is the point to win the
battle which never can be won. Not because the linux/funtoo desktop
will never be like... or state of the art.
I do lots of work for other people managing them struggle their
battles with desktops. Mac and Win. And i do have another point of
view because i also be psychologist and can see why they struggle.
They just stay with what the had forever.
They dont want to learn.
They too (excuse me) stupid to see the equivalence of different
systems because there mind is just frightend that it is different,
because they told so.
They say whow to all the possibilities a programm can offer, but they
use just the basics.
Even on a mac they can't find any logic aside clicking a icon. So
where is that intuitiv desktop?
They just buy another new PC because the registriy is blown and
slowing down the whole system. But they say its because the system is
really old and new technologie and the new OS-version will solve all
problems. Bacause the told that it is smart and quick and funny and
more safe. And than everything starts from new for another two to four
years.

So tell me something about itunes what rhytembox can't do? This is
just one for all of them. Tell me one program linux cant do at least
same.
What is really the shiny beauty of a windows desktop?
We dont start with talking about easy updates. They just are'nt done
or done but, errors not seen. Shall i tell you about my results after
searching there systems for viruses, trojan... (and yes - they have
mostly more than one inti-X on there system)
-
So if a user is willing to spend a bit of time to understand what he
is doing or what the system does, he whould - like me - stay happy
with the damm quick fluxbox and has everything he needs.
If he just wants to be a mouse-ruled user and likes blinking here and
there, sliding-fading and .. (all what is not needed) then you can be
shure - the never change to linux and linux should not look after
them. What for?
Even the point of criminalisation because cracked programms can't
force them to switch. Once more - they do not understand.
And really Ubuntu is just not the best example. What is good or better
there?
Technically Ubuntu for me is a horrortrip. Did you ever count the
started services after a fresh minimal install? What is the difference
to Windows?
The difficulties do not start because the linux-desktop-system is not
well made, it's because it is impossible to serve couch potatoes!

Go your way and do not bother desktop more then it is already managed.
cheers
Karl

[Daniel Robbins]

I actually think we can do a great deal with Linux on the Desktop, but first we have to give up with the existing strategy.

This is what Ubuntu has started to do with Unity. While their implementation is not ideal, they do have more control this way.

We can't repeat the same failing strategy and expect to "win".

-Daniel

[Daniel Robbins]

Nick,

I think you have some insightful points here and we want to work in this direction. There is no reason why we can't build AMIs with Metro.

Nick, I think there is a great deal you can do to help us as a sysadmin :) Are you volunteering? If so, I say yes :)

-Daniel

[Nick Gotsinas]

Daniel, yes, like I said willing and available :)

--

[PerfMonk]

You forgot the other part of the presentation :

Why Linux does not sucks!!! not even a little bit!

[url]http://www.youtube.com/watch?v=BfLqLK7VdQY&feature=relmfu[/url]

And I cite Brian Lunduke the presentator (this is not my words):
"Anyone who want to make that case (that Linux sucks) is a moron!"

And he even talk about reasons why "dumb people say linux
sucks" (still his words).

Everybody that was brainwashed to think that Linux sucks should
listen to this presentation.

Regards,

Bernard :=)




On 23 mai, 13:19, Adam Smith <absmi...@gmail.com> wrote:

[UNIXgod]

On Wed, May 23, 2012 at 7:21 PM, PerfMonk <tremblay...@gmail.com> wrote:

> Everybody that was brainwashed to think that Linux sucks should
> listen to this presentation.

Who's on this list/usenet group brainwashed?

I think it's pretty sane to consider a stable production worthy server
first and desktop to hack around with second. Basically the same
strategy is implied in BSD servers.

I also imagine once it's been around the block and is considered
"safe" for production the adoption rate will rise.

My two bits anyways

~

[Sylvain Alain]

Hi everyone, I have a couples of ideas before the Funtoo profiles are up and running.

Since the focus will be on Server side and mostly on the Funtoo profile, I would like to know how many efforts is required to keep up with the upstream(Gentoo) for the Desktop stuff(Openrc,udev,consolekit and others....) ?

Also, how different the custom Funtoo ebuilds are compare to the Gentoo ebuilds  ?

Would it be possible to actually bump the Gentoo version that are higher then the Funtoo ones and to hardmask them ?

With that idea in mind, all the Funtoo Server stuff will be protect and the Desktop Fanatics will be able to update their box.

Finally,is it possible to switch back and forward from Gentoo to Funtoo on these packages(udev,openrc). I don't talk about boot-update since it's really a must for the Funtoo users.

My main concern is that I hope that we still be able to update our box even it's not the priority of Funtoo right now.

2012/5/23 UNIXgod <uni...@rubyprogrammer.net>

~

--
To manage your subscription, visit this group at
http://groups.google.com/group/funtoo-dev?hl=en
---
Also be sure to check out:
 Funtoo Forums: http://forums.funtoo.org
 Planet Larry: http://larrythecow.org

--
Salut
alp
Sylvain

[Marc 'Judge' Richter]

One has to focus on a workflow and choose a fiting tool matching the needs. Choosing a (popular) tool and pressing a workflow in there is not a good or comfortable way of bringing a new system in.

If Redmine isn't what you thought it would work out, then let it go before you invested too much work in it. That's my experience ...

But IF there is a tool, it is difficult to handle when nobody really knows "THE" way to work with it. The more newbee you are, the more difficult it is to diff between right and wrong.

I have absolutely no preference in the tool to use! But I have the very strong wish that when a tool has to be used by the project a documentation has to be publically available which clearly explains when and how this is to be used. Perhaps I miss something here, but I haven't seen this to be done with the Redmine launch ... and I think much of the points you raise are happening because much people simply don't know better.

Am Mittwoch, 23. Mai 2012 16:07:51 UTC+2 schrieb Daniel Robbins:

Well, I don't know if there is fantastic documentation on Redmine, and I am personally struggling with it. I may switch to a free version of JIRA as I think it could work better for us.

I see the same types of problems with bugs that you see, and ideally the bug tracker would help us with the workflow a bit more.

-Daniel

-Daniel

[PerfMonk]

OK Linux desktop (according to wikipedia) is about 1,66% of the
desktop market.

But my question is not answered. How many people here use Funtoo on
server ? How many use it for desktop ?

I think the userbase is mostly desktop and the server is marginal.

So why cut off the base to goto a server distro and dish your best
clients ???

I just want to understand,

Regards,

Bernard

On 23 mai, 12:56, Daniel Robbins <drobb...@funtoo.org> wrote:
> Facts:
>

> Linux has 1% desktop market share. (http://www.omgubuntu.co.uk/2012/01/is-linux-marketshare-on-the-rise-i...)

> - even after ~8 years of Ubuntu.
> Linux powers Amazon, Google, etc. (millions of servers)
> Linux powers 63% of all Web sites.
> Linux powers basically all supercomputers.
> Linux powers all Android mobile devices (up to 72% smartphone marketshare
> in parts of Europe.)
>
> We are not giving up on desktop but we must have priorities and the
> priorities are "server" (ie. cloud, Web, etc.), followed by
> mobile/embedded, and desktop is more of a strategic target which we are
> starting.
>
> If the Linux desktop is a success, I do not think it will be running KDE
> and GNOME.
>
> But in order to "win" the desktop, we must live in reality first. And then
> we can make honest decisions.
>
> For KDE, GNOME, and other DEs/WMs, we will make the experience better. We
> will be doing some stuff that no one else is doing on the desktop, which
> will be innovative.
>
> The first step is to deal with reality. If the Linux desktop is so good,
> why aren't more people using it? The reason is, even though you may like
> it, it is not competitive with MacOS or Windows in the larger world for a
> number of reasons.
>
> -Daniel
>

[Marc 'Judge' Richter]

Hi PerfMonk,

no need to discuss this here: Daniel made a decision as BDFL, even explained why, and servers get the current focus. No need to discuss if everybody likes this or not.

And Desktops won't be not supported anymore - just the current focus is set to servers to reach a specific goal and THEN we look where to go from there.

Best regards,

Marc

[PerfMonk]

Thanks for the precisions!

But let's say it's a little bit disturbing for most people concerned
with the desktop.

regards,

Bernard :=)

On 24 mai, 10:28, Marc 'Judge' Richter <marc.richter.

> ...
>
> plus de détails »

[Daniel Robbins]

PerfMonk,

I appreciate your contributions to the discussion.

Let me explain my position. I make decisions based on potential in the overall market, rather than based on what our current user base is doing.

I think this is the better way to make decisions, since you are exploring the absolute potential of server, desktop, etc. rather than basing decisions on whoever happens to be using Funtoo right now.

For example, I tried to get Gentoo to make binary packages. The majority of developers didn't want it, following a similar philosophy ("we all don't care about binary packages!") but where could Gentoo be now if it had binary packages? The larger distro ecosystem was more binary-based and wanted a binary solution. But if you take a vote from Gentoo devs, they would overwhelmingly say "no". They were afraid "from source" would go away, but we can do both.

Back to desktop. I know most/all of the people using Funtoo are the most hardcore of the hardcore Linux users, so I am not surprised that many use Linux on the desktop. But they are not representative of the overall Linux world or computer world. We are special :)

I was hoping to shake up people a bit and get them to realize that (to use the popular term) they are the 1% and the 99% will not touch Linux because it's a waste of time for them and all they want is to get stuff done quickly and run programs they buy. So we need to balance our enthusiasm with a bit of reality. We can be personally excited about Desktop but this decision (server vs. desktop) is not based on our personal opinion but on market realities.

However, with the funtoo-1.0 profile, this can start to change.

So, regarding market share -- right now, all the desktop stuff on Linux is grabbing 1-2% market share, which is pretty bad. Frankly there is not much difference between 2% and 0%. The desktop is a great thing to work on but it is not worth the effort to fight for a piece of the 2% when we can fight for a piece of much bigger and more exciting areas for Linux right now.

Because the desktop market share for Linux sucks, to really make a difference we will need a *brand new* approach for desktop, very innovative. I want to do it, but I cannot launch it right this second. A brand new disruptive approach will allow rapid growth and for us to reach new people who will not touch Linux right now. We cannot fight for the 2%, we need to be looking at 10% opportunities and up.

Hope this clarifies,

Daniel

[Daniel Robbins]

I agree about switching to another tool. I may switch to JIRA and get a free license. We use it at work, and we have a nice workflow set up. The Redmine interface isn't really working the way I want and there are benefits to a commercially-supported product -- better documentation, fewer quirks with product features, etc.

I will look into switching soon. 

I think Redmine has too much of a learning curve and I don't really have time to invest in our bug tracker. I need to read a few FAQs and deployment guides, tweak for a few hours, and be in business.

My feeling is that Redmine is really capable but it requires a lot of work to get it working and to learn the "redmine way" of doing things.

-Daniel

[Ryan Harris]

Well I guess the original post isn't too bad. You know, once you get past the part telling us how much our desktop sucks, how much better apple and windows is, how they hire "real" developers (not sure what you'd call people who work on Linux desktop then). Unfortunately unlike some people, not all of us can afford a copy of Windows Server or a Mac to use as a desktop so for those people they are just stuck with this crappy stuff.

Anyway, once you get past that, the rest of it makes pretty valid points and the biggest point is that we most certainly can do both. The biggest thing right now if your a hardcore desktop user (which i bet most of Funtoo is) is to have patience. Funtoo has always prioritized servers so we can't break everyones stuff just to get newest desktop stuff out right now. You might run into weird blocks and stuff right now but just bear with us for a little while as we figure out the best way to do this and get the Funtoo profiles out which is vital. The best solutions are usually not the fastest ones.

[Daniel Robbins]

I admit is was a pretty harsh email, but that was kind of the point. I was frustrated, so you could look of it as sort of (shock?) therapy, for both myself as well as those who read it, which I felt was needed in this case.

My point was not about "real" developers, but "real" UI research and design, which *does* take money. It takes a lot of time, and it requires a lot of focus groups and other real-world testing to vet the UI and ensure that it actually works for people. 

Bigger companies will actually test their UI on real people and get scientific usability data back from these sessions which they will use to further refine their UI. In contrast, Open Source pretty much as a rule does not do this and the UI design is whatever someone cooked up, and is often not optimized for the general public. Often a developer is doing triple-duty as a UI designer and graphic artist and the results are often not ideal, if not downright gross. Often the focus is on new features rather than optimizing the experience.

And I agree with your summary. It's critical that we don't try to do one thing at the expense of the other, because this is a lose-lose strategy.

That being said, I think we can do great things on the desktop and I actually have a plan to do so. But first we need to get a handle on maintenance and server stuff.

-Daniel

On Thu, May 24, 2012 at 5:52 PM, Ryan Harris <r...@funtoo.org> wrote:

Well I guess the original post isn't too bad. You know, once you get past the part telling us how much our desktop sucks, how much better apple and windows is, how they hire "real" developers (not sure what you'd call people who work on Linux desktop then). Unfortunately unlike some people, not all of us can afford a copy of Windows Server or a Mac to use as a desktop so for those people they are just stuck with this crappy stuff.

Anyway, once you get past that, the rest of it makes pretty valid points and the biggest point is that we most certainly can do both. The biggest thing right now if your a hardcore desktop user (which i bet most of Funtoo is) is to have patience. Funtoo has always prioritized servers so we can't break everyones stuff just to get newest desktop stuff out right now. You might run into weird blocks and stuff right now but just bear with us for a little while as we figure out the best way to do this and get the Funtoo profiles out which is vital. The best solutions are usually not the fastest ones.

[PerfMonk]

Thanks a lot for the answer. I feel better now by knowing the desktop
will not be forgotten and put aside. I agree that the server is a
strategic ground for Linux. The idea of a desktop/server profile may
be an interesting way to meet both needs : stability and state-of-the-
art.

I feel better also knowing that there is of new desktop approach
taking form even if it is still on the drawing board for now.

Regards,
Bernard

> ...
>
> plus de détails »